Serving the Quantitative Finance Community

Search found 7 matches

by albanese
October 29th, 2007, 8:53 pm
Forum: Numerical Methods Forum
Topic: Geman & Roncoroni Jump Processes
Replies: 8
Views: 104582

Geman & Roncoroni Jump Processes

<t>> This is exactly what concerns me. When you look at the history of computing the languages and tools which have stood the test of time are those that are 1) > ISO/ANSI/IEEE standard (Cobol, Fortran, C++) and/or 2) have evolved from previous, immature technologies (for example, MPI, OpenMP)I woul...
by albanese
October 16th, 2007, 12:38 pm
Forum: Numerical Methods Forum
Topic: Geman & Roncoroni Jump Processes
Replies: 8
Views: 104582

Geman & Roncoroni Jump Processes

<t>> How can we get guaranteed error bounds in general? Are you using some interval arithmetic?Very good question. The answer is no, that's why this is subtle. If you use interval arithmetics you arrive to worst case error estimates which unavoidably tell you errors explode exponentially in the numb...
by albanese
October 14th, 2007, 11:48 am
Forum: Numerical Methods Forum
Topic: Distributed Finite Difference
Replies: 9
Views: 66386

Distributed Finite Difference

<t>Traditional parallelism is achieved with clusters/MPI but the emerging new trend is to use massively parallel GPUs and manufacturer provided BLAS. The two strategies are quite different. With clusters you try to use sparse matrices without storing them and apply them or invert them on vectors. Na...
by albanese
October 11th, 2007, 5:11 pm
Forum: Numerical Methods Forum
Topic: Geman & Roncoroni Jump Processes
Replies: 8
Views: 104582

Geman & Roncoroni Jump Processes

<t>> I read somewhere that Jack Dongarra claimed speed up of 10 even with LU and for eigenvalue problems. It seems almost too good to be true but it might be worth a try.Dongarra used iterative methods in mixed precision, not purely in single precision as I suggest. Dongarra's idea is to implement i...
by albanese
October 10th, 2007, 10:01 am
Forum: Technical Forum
Topic: Albanese and Trovato's Stoch Vol Model for Callables
Replies: 6
Views: 121339

Albanese and Trovato's Stoch Vol Model for Callables

<r>Hello everyone, I think the point Manlio and I want to make is not that of proposing yet another short rate model. As a matter of fact we wrote several papers recently, each one with a new model specification. Our main objective is rather to develop a viable numerical framework for models specifi...
by albanese
October 9th, 2007, 4:43 pm
Forum: Numerical Methods Forum
Topic: Geman & Roncoroni Jump Processes
Replies: 8
Views: 104582

Geman & Roncoroni Jump Processes

<r>Actually MATLAB is a good platform for the implementation we propose. The keypoint there is that one needs to have access to an efficient implementation of a matrix-matrix multiplication algorithm. Matlab does that for you.I'd be interested to know how come you find the implementation description...
by albanese
May 15th, 2007, 1:12 pm
Forum: Technical Forum
Topic: Structuring a CMS spread note
Replies: 2
Views: 73524

Structuring a CMS spread note

<r>What one needs is to price the CMS spreads structure by modeling directly and hopefully correctly the monetary policy scenarios. I just wrote a note on this and posted it on my web site at <URL url="http://www.level3finance.com/publications.html">http://www.level3finance.com/publications.html</UR...